{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import datetime\n",
    "from datetime import date\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.formula.api as smf\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "import matplotlib.font_manager as font_manager\n",
    "mpl.rcParams['font.family']='serif'\n",
    "cmfont = font_manager.FontProperties(fname=mpl.get_data_path() + '/fonts/ttf/cmr10.ttf')\n",
    "mpl.rcParams['font.serif']=cmfont.get_name()\n",
    "mpl.rcParams['mathtext.fontset']='cm'\n",
    "mpl.rcParams['axes.unicode_minus']=False\n",
    "import pystan\n",
    "import pystan.experimental\n",
    "import arviz as az\n",
    "import researchpy as rp\n",
    "from scipy import stats\n",
    "from scipy.stats import ttest_ind\n",
    "#%load_ext rpy2.ipython"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%run create_data.py"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Manuscript Graphs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Figure 1:  Pct. of Minority Voters by City/County"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "e=pd.read_csv('data_figure_1.csv')\n",
    "sns.set(context='paper', style='white', palette='dark', font_scale=1, rc={'lines.linewidth':.70})\n",
    "sns.barplot(x='city_county_name', y='non_lat_electorate_cc', data=e)\n",
    "plt.xticks(rotation='vertical')\n",
    "plt.ylabel('Pct. Minority Voters')\n",
    "plt.xlabel('Locality')\n",
    "plt.xticks([])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 2 Avg. Pct. Minority Candidates by Party List"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEACAYAAAByG0uxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/d3fzzAAAACXBIWXMAAAsTAAALEwEAmpwYAAAff0lEQVR4nO3de1xUdf4/8NcwF2a42GAITBdXa1es/GYuJpFojzQvKUtcVLalxfx6yRtqIphQKATiZqAbpUuhJqbrFUSwX5r4LZU182t2MU13xRLY70AGCMRlmGF+f/hlvlLeOMzMOTO8nn8hzJzzPlDnNZ/POefzlpnNZjOIiIgEcBG7ACIiclwMESIiEowhQkREgjFEiIhIMIYIEREJxhAhIiLBFGIXYE/+/v5il0BE5JDOnz9/w+/3qBABbv6LICKiG7vVB/AeFyI9QXtLK8wmk9hlEDk8mVwOF7Wr2GVIGkPECZlNJtTk5YtdBpHD6x0TIXYJkscL60REJBhDhIiIBGOIEBGRYJK5JnL8+HHs3LkTXl5e6N+/P06ePAlvb28MGjQIw4YNw6JFi/DYY4/h559/xksvvQSFQoFFixbB398fHh4eWLZsmdiHQETU40gmREpKSpCQkACdTodNmzYhJiYGQ4cOBQBUVFTgqaeeQmxsLBobG5Geno558+ZZvhcXFydy9UR0vfrWFrQY28Quo9ta9XrI3TRil2EVGo0GWq3W6tuVTIjMnDkTGzZsQH19PaKionD48GHs2bMHY8aMwYABAyyv8/DwsHx99OhRXLlyBcOGDROjZCK6AYPJiNgPd6HVZBS7lO77f7vFrsBq1Go1CgoKoFKprLpdyYSIr68vEhMTYTQasXTpUmRmZgIAFi5ciPj4eMvrGhsbLV+PGDECsbGxdq+ViG5OJVcge8JkpxiJ3BU+zqlGItYOEEBCIbJr1y6cO3cOBoMBwcHBWLFiBWQyGYYMGQIA+PTTT3H16lXLNREikq5ermr0clWLXUa39fbzg9zdTewyJE3Wk9rj+vv794hlT0w/N/FhQyIr6B0TwRDBrc+dvMWXiIgEY4gQEZFgDBEiIhJMMhfWyXpkcjkXjiOyAplcLnYJkscQcUJcupqI7IXTWUREJBhDhIiIBON0lhMytTTA7ARPC5NzkimUkKs9xS6DrIQh4oTMxjb8e+MsscsguqF7/vNdsUsgK+J0FhERCcYQISIiwRgiREQkmNWuiVizM2F5eTnWr18PtVoNo9GIGTNmoLCwEOHh4SgvL0dJSQlGjx6NnJwc+Pn5oX///lzZl4hIBFYLEWt2JszJyUFiYqKlAZXJZAJwrQnV5cuXkZSUhM8//xyhoaGIiIjA0qVLrXUYRETUBVYLEWt2JjSZTPDw8MCZM2ewdetWjBs3DgCQl5eHbdu2QSaTAQCKi4tRUFCA0NBQax0GERF1gdWuiXR0JkxLS8MHH3yAuLg4ZGRkoLCwsNPrftmZMCUlBVFRUSgpKUF6ejqqqqogl8vR2NiIQYMGISwsDDU1NQCAlJQUJCcno6GhAQAQEhKC3NxcnDx50lqHQUREXWC1kUh3OxOOHj0ao0ePBgC89NJLSE9Ph7u7O1pbW/H888+jsrIS99xzD+Lj45GUlITw8HAAgKurKx577DEcO3YMwcHB1jocIiK6A+xs6ISMjTV82JAk657/fBcKj95il0FdwM6GRERkEwwRIiISjCFCRESCcQFGJyRTKLnIHUmWTKEUuwSyIoaIE+Iy20RkL5zOIiIiwRgiREQkGKeznJChpQHtJoPYZdB1XOQqqDjNSE6IIeKE2k0GfLr9z2KXQdd56o9bxC6ByCY4nUVERIIxRIiISDCGCBERCSbpayL5+fnYu3cv8vLykJ+fD6VSiaNHj8Ld3R2hoaG4dOkS7r33XgQGBopdKhFRjyTpEAEAb29vHD9+HADwxhtv4ODBg9BoNACAS5cuiVlaj9DUYobBKHYVjk+vr4arG++Y6y6NRgOtVit2GXQdyYdIZGQk8vPzERQUhNmzZ2P16tVobm7GzJkzxS7N6RlNZuQUGtBmErsSx/fuvvlil+AU1Go1CgoKoFKpxC6F/pfkQ0ShUGDYsGEoLS3FlClTEB0djStXrmD9+vV45JFHxC7PqSnkMrz0nIojESsIDMmCq5tW7DIcnkajYYBIjORDBADCw8ORm5uLgQMH4uDBg2hsbER4eDj+/e9/Y9OmTdi/fz8CAgLw3HPPiV2q03FTy+AmdhFOwM/PB2r3u8Uug8jqJB0iERERlq8PHDhw29cQEZF98RZfIiISjCFCRESCMUSIiEgwhggREQkm6QvrJIyLXMVVYyXGRc7bUsk5MUScEPtWEJG9cDqLiIgE40jECbW01sNkahO7jB5PLldC7dpL7DKIbIoh4oRMpjZ8sOdPYpfR470QuU3sEohsjtNZREQkGEOEiIgEY4gQEZFgDndNJD8/39LNcPbs2dDpdKiqqsLy5cvh6+srdnlERD2Kw4XI9bRaLZYvX46PP/4Yp0+fxvjx48UuSfJaWsww8sYtu9Drq+GmYTdDe2DHQ/E4dIjU1dUhPT0dRqMRS5YsEbscyTMZzdi/2wQTm0zZxYd72M3QXtjxUDwOHSJarRZJSUlil+Ew5AoZJk6ScyRiJ2Hj/wo3jVbsMnoEdjwUj0OGSEc3w3379mHVqlVil+NQ1GoZoBa7ip7Bz88H7m7sZkjOzeFCJCIiwtLNMDU1VeRqiIh6Nt7iS0REgjFEiIhIMIYIEREJxhAhIiLBHO7COt2eXK7kCrISIJcrxS6ByOYYIk6IPSyIyF44nUVERIJxJOKEGg31aHOAzoZKuRIeKo6aiBwZQ8QJtZnasPijaLHLuK2s8VvFLoGIuonTWUREJBhDhIiIBGOIEBGRYJK+JtLRxXDIkCFIS0sDAPz0009QKpWYO3cu3n77bXh7e2PQoEGWRRmJiMh+JB0iHVQqFVJTU9HW1oalS5ciMTERRUVFiImJwdChQ8Uuj4iox3KIEOmQlZWFqVOnwtvbG5GRkXjvvfewZ88ejBkzBqNGjbLpvuvq6tDc3GzTfVjL1ZY6GOvNYpdxW1X6arSo2T7WmtgmluzNYUKksLAQ/fr1w+DBgwEAvXr1QlxcHABg4cKFNg0Rg8GA6OhotLS02GwfPdG8PLaPtTa2iSV7k3yIbNq0CUVFRdi9ezemTJmC5ORk+Pr6wt/fH8eOHYNMJsOQIUNsWoNKpcLWrVsdaiSSfmSR2GXcVtLItbhLrRW7DKfCNrFkbzKz2Sz9eQ8r8ff3x/nz58Uuw+Zqm39ymIcNvTRsH0skdbc6d/IWXyIiEowhQkREgjFEiIhIMIYIEREJJvm7s6jrlHKlQ6yQq2TnPyKHxxBxQuzRQUT2wuksIiISjCMRJ9RgaIbBZBJl3yq5HJ4qjSj7JiL7Y4g4IYPJhJiP1ouy77zxc0TZLxGJg9NZREQkGEOEiIgEY4gQEZFgkr0msmPHDnz77bcAgJ07d2LSpElwcXHB5cuX8cgjjyA4OBg7d+6El5cXgoODbd5PhIiIfk2yIRIVFQUA2L59O9544w2EhoaisbERSUlJmD9/PjIzM5GQkACdTidypUREPZdkQwQAvvzyS1y+fBkJCQkwm81IT0/HkiVLoNFoMHPmTGzYsAH19fWIioqyeU8RqbiTDot1LU1or2+yU0WdVeurYFA3iLLvO8HOf0TWJdkQuXLlCvLy8vDGG28AAHJycjB+/Hjcf//9AABfX18kJibCaDRi6dKlPSJEHKHD4vy8I2KXcEvs/EdkXZJtSrVo0SK0tbXh7rvvxuXLl/HZZ59hypQpAK41SFGpVDh37hwMBgOeeOIJhISE3HabztCU6k5HIouPfGCnijrLGvkCtGo3UfZ9JzgSIeq6W507JTsSWbt2rdglSJJWq73tSVDV3AiXXuKcyH38fHG3xkOUfROR/fEWXyIiEowhQkREgjFEiIhIMIYIEREJJtkL6yScSi4XbTVdlVwuyn6JSBwMESfEfh5EZC+cziIiIsEYIkREJBins5xQQ2srDO0itcd1kcPT1VWUfROR/TFEnJCh3YSp+wtE2ffmieGi7JeIxMHpLCIiEowhQkREgjnEdFZ+fj4KCgqQkZGB8vJylJSUYPTo0cjJyYFOp0P//v0xa9YsscskIupxHGokcvToURw5cgRJSUlwcXFBaGgoMjIyHH55dyIiR+UQI5EOeXl52LZtG2QyGQCguLgYp0+fxvjx40WuzDGZm5thbjNadZvVej0Maus97Mj+H0TS5lAhkpKSguTkZKxcuRIAEBISgoiICJGrckxmoxEtH+wCjNYNkflbd1l1e+xESCRtDhEiRqMRSqUS99xzD+Lj45GUlITwcN5K2h0yhQLqFyZbfSSS+fRYaK08EmGAEEmX5EPkq6++wieffILs7Gy4u7sDAN566y2Rq3IOMo0GMisvs+Xj54e7NdJtj0tE1iX5EBk8eDDWrVsndhlERHQDDnV3FhERSQtDhIiIBGOIEBGRYJK/JkJdp3KRi7YQosqFnQ2JehKGiBPiUuxEZC+cziIiIsEYIkREJBins5xQQ2sbDO3tNt+PysUFnq5Km++HiKSLIeKEDO3tmL7/mM33s2FisM33QUTSxuksIiISjCFCRESCMUSIiEgwq4bIggULYDKZAAC7du0SvFz7K6+8AgDYv38/1qxZg7KyMmRnZ1utTiIisg6rXlh/8skn8Y9//AMjRozAsWPH8OCDD8JoNGLNmjVobW3F3XffjTlz5iAiIgITJ06EXq9HUlIStmzZgu+//x5NTU1ITEzEuXPnkJ2djcjISBw9ehQqlQqff/45srKyYDKZEB8fj82bN6OiogJ9+vSRfH/1uro6NDc3229/LQa019fZfD/Xuhg6X68PdlMkunNWDZEJEybgL3/5Cx544AHodDrU1dWhtLQU1dXV6Nu3Ly5dugQAuP/++zF9+nQsW7YMAPD1119j9erV+PDDD3Hy5Ek89NBDiI2N7bTtAQMGYPHixUhLS0NDQwPMZjPc3d1RUlIi6RAxGAyIjo5GS0uL2KVY3fytYldgG+ymSHTnrBoivXr1Qnt7O7Zv347Q0FDk5eXBZDIhKCioUxtbjeZaJySz2XzD7XT0UL+Z+vp6VFRU4NVXX8W3335rvQOwAZVKha1bt9p9JLL0v/7b5vv5y9NDoXXSkQgDhOjOWP05kXHjxmHt2rWIi4sDAAQHB2PFihX45z//CYVCYfn+9QYNGoT09HQ0NjYiJSUF1dXVSEtLw7Rp0yyvuXDhAjIzM6FWq+Hn54fa2lrk5uaiqqrK2odgdVqt1q7TI6rmVrj0sv3+rnUx5DpdRD2ZzHyz4YAT8vf3x/nz58Uuw+Z+am6128OGDBEi53ercydv8SUiIsEYIkREJBhDhIiIBOMCjE5I5eJil8URVS78DELU0zFEnBCXZycie+FHSSIiEowhQkREgnE6ywn9bDDBaLLt4z8KuQzuKrlN90FE0scQcUJGkxmvffQ/Nt3H6+N1Nt0+ETkGTmcREZFgDBEiIhKMIUJERIKJEiL5+fk4ceIEAGD27NlISUnB7NmzodfrAQCffPIJ/vjHP/7qfR0dD68XExODAwcO2LZgIiK6IdFHIlqtFsuXL0dkZCTOnDkDADh8+DDGjBmDs2fP3vK9//rXvxAcHIyjR4/ao1QiIvoF0UOkrq4Oy5cvx7vvvougoCDU19dDoVAgNDQURUVFt3xvYWEhQkJCoNFoUFNTY6eKiYiog+i3+Gq1WqSkpODvf/87vv76a/zwww+orKxEdnY2vvnmG3z33XfYs2cPxowZ0+l97e3tKC0txdWrV/Hjjz+iuLgYMTExIh0FEVHPJEpTqvz8fBw8eBA+Pj7YvXs3zp49C4PBgISEBLS3tyMrKwsKhQK7du2Ct7c3nn76aQDXrol0tC2977774ObmhhdeeAFmsxkLFixAdnb2LffbU5pSXW022uU5kbs0on8GISI7uNW5k50NnRBDhIisiZ0NiYjIJhgiREQkGEOEiIgE46S2E1LIZTZfIFEhl9l0+0TkGBgiTohLtBORvXA6i4iIBGOIEBGRYJzOckKG1na0t3f/8R8XFxlUrvycQUQ3xxBxQu3tZvzXvtpub+fpUC8rVENEzowfM4mISDCGCBERCcYQISIiwSQdIkI7IBIRkX04zIX1jg6IH3/8Mc6cOQM/P79OHRAffvhhsUsUVV1dHZqbmwEArS0m1NXXd3uben0LXNXWe3BRo9FAq9VabXtEJD6HCZGODohnz57F+++/36kD4saNG3t0iBgMBkRHR6OlpcWq283dbtXNQa1Wo6CgwNIThogcn8OEyO06IBqNRigUDnM4VqVSqbB169ZOI5Hjh7o/Egl6ppfVRyIMECLnIvmz7qZNm7B//37s27cPq1atQmRkpKUD4jvvvGPpgHj06FFLB8SeSKvVWqaKWppN0PZSd3ubfn5eUGu4DhcR3Rw7GzqhlmaT1R42ZIgQETsbEhGRTTBEiIhIMIYIEREJJvkL69R1Li4yqyye6OLC7oVEdGsMESfE5duJyF54tiEiIsEYIkREJBins5yQqbkdZlPXH/+RyWWQa/i5gojuHEPECZlNZlTk/tjl9903o48NqiEiZ8aPnUREJBhDhIiIBGOIEBGRYJK+JpKfn497770XQ4YMQVpaGgDgp59+glKpxPPPP4/169dDp9Ohf//+mDVrlsjVEhH1PJIOkQ4qlQqpqaloa2vD0qVLkZiYiIsXLyI0NBQRERGIi4sTu8Q7cn33QVsyNZnwY1NNl9+nqDJC3iDdVXvZGZFIehwiRDpkZWVh6tSp8Pb2xsWLF1FcXIzTp09j/PjxYpd2W7bqPmhVJWIXcGvsjEgkPQ4TIoWFhejXrx8GDx5s+V5ISAgiIiJErOrO/bL7oC2Zmkz4nx1dH4noonpD7ibtkQgDhEhaJB8imzZtQlFREXbv3o0pU6YgOTkZvr6+GDp0qNilddn13QdtydhogtGt639aP98+UHhIN0SISHrY2dAJGRtNgh82ZIgQ0S+xsyEREdkEQ4SIiARjiBARkWAMESIiEkzyd2dR18nkMkEr8srkbIdLRF3DEHFC7AlCRPbCsw0REQnGkYgTam9pg9nYtcd/ZAoZXNRKG1VERM6KIeKEzEYzrmw41aX3eE8PsFE1ROTMOJ1FRESCMUSIiEgwhggREQnmMCGSn5+PmJgYy9cnTpzA1q1bHaYhFRGRM3KoC+ve3t44fvy45d9nzpyBn58fampq0Lt3bxEr6x5rdzw0NbWh9ueu9RNpq9JD3iDu3VnsXEjkeBwqRCIjI5Gfn4+goCCUlZXh/vvvR3BwMIqKijB16lSxyxNEMh0PJdDVkJ0LiRyPQ4WIQqHAsGHDUFpaisrKSuh0Ouj1ely6dMlhQ8QWHQ9NTW2o3fFNl97jFfUfkLuJPxJhgBA5FocKEQAIDw9Hbm4uamtrsX37dgDAX//6V1y4cAEDBgwQuTphrN3x0NRogNK9skvv8fb1g9yDJ3Ai6hqHCZHre6kfOHCg088WLlxo73KIiAgOdHcWERFJD0OEiIgEY4gQEZFgDBEiIhLMYS6s052TKWRdXpVXpmBXQyLqOoaIE2JfECKylx4XIv7+/mKXQETkNGRms7lrLfCIiIj+Fy+sExGRYAwRIiISjCFCRESCMUSIiEgwhggREQnGECEiIsF63HMiDQ0NSE9Ph0wmg7+/P1588UXs3bsXp0+fhtlsxooVK+Di4pjZqtfrkZmZCQCYPn06Bg4cKHJFwh0+fBiHDx9GfX09YmNjsWXLFgDAkCFDEB4eLnJ1wtXU1CAqKgobN27Ee++9B8Cxj8lkMuGtt95CY2MjAgMDcezYMQCOfUyVlZVYuXIlPD09ERAQgG++udbgzVGP6cKFC1i3bh0efPBBTJs2DWlpaZDJZAgJCcHw4cOxevVqtLS0wM/PDzNnzuzy9h3zbNkNnp6eWLVqFTIyMnD27FkAwGeffYaUlBQ8+uijOHXqlMgVCrd7924sWLAAr7/+uqVhl6MaNWoU0tLSMHfuXBw4cACDBw9GamoqTpw4IXZp3bJ582Y888wz0Ov1TnFMJSUlqKmpgYuLC7y8vJzimC5duoTw8HCsXLkSRUVFDn9MAwYMwJIlSwAAhw4dQlhYGDIyMlBcXIzKykq4urritddeQ3l5OQwGQ5e33+NCpMOhQ4fw+OOPAwBksmvrRul0OlRVVYlZVrdUV1dDp9NBrVajtbVV7HK6rb29Hdu2bUPfvn2h0+kA/N/fyhF9+umn+P3vfw83Nzd88cUXTnFM33//PQIDA/HKK6/gxRdfdIpjevjhh7F582b86U9/wqRJk5zimDpUVVVZjgf4v3MGAHh5eaGurq7L2+xR01n5+fk4d+4cAgMDcf78ecybN6/Tz/V6Pfr27StSdd3n4+MDvV6PPn36wNXVVexyusVsNmPVqlWYMmUKmpubUV5eLnZJ3Xb69Gn8/PPP+Oqrr+Dq6oo+ffqIXVK3+fj4wNPTE3K5HAEBAdDr9WKX1G0FBQVYtmwZ/P39MW3aNISFhYldktX4+vpCr9fjN7/5DYBrf7+OKcja2lpBbbp73LInV65cQVhYGEaNGgVPT0/Ex8dj7969+Prrr2E0Gh3+msiaNWsgk8nw4osvOvQ1kc2bN6O4uBgPPfQQgoODUVpaCrlcjkcffdTh/6fOzs5GWFgYcnNzHf6YmpqasHLlSsjlcgQGBuLEiRMOf0zfffcd/va3v0Gj0ViuiTjyMen1eqxevRqXLl1CTEyM5W/07LPPYvjw4XjzzTdhMBjg4+ODGTNmdHn7PS5EiIjIehzzIzcREUkCQ4SIiARjiBARkWAMESIiEowhQpJRXl6OJ554olvP6hw8eBBZWVl4++23AQAbN25EdXV1p9dUVFTg0UcfxeXLlwEAmZmZmDt3LgBg5cqVXd5nVVUVcnNzUVFRgfT09Nu+/sKFC0hISEBqaioSExO7fFvsiRMn8P777+Obb77Bvn37Ov3slVdeQX19/U3fY0u/3PfNfpcdv2tyDj3qORGStp07dyI5ORm7du3CzJkzsWzZMmRlZeHChQvYv38/Ro4caXn48PTp0zc8KR4/fhyLFy/GihUrcPbsWfj4+MDHx+dXr3v66aeRn5+POXPmoLm52fL9iooKAMCECRMwefJkfPnll0hNTcUPP/yALVu2QKFQYOzYsfjd736HuLg4jB49Gs8++yyqqqpw6tQpfPXVV9iwYQOUSiX8/f0RGBiIJUuWIDU1FW5ubgCAtWvX4s0337T822AwoLi4GOfOnUNNTQ2WLVuGQ4cO4cSJE+jXrx+ampoQFxeHNWvWwGAwoLa2FgMHDkRTUxNqampQU1ODlJQU9OvXDxcuXABw7cHGkydPoqamBnPmzMGpU6dw6tQpuLu7Y+TIkZYlVzw9PTF37lwkJCTAz88P/v7+nW5jXbduHWpqalBbW4vk5GTk5eWhoaEBGo0GvXv3xtSpU2/69+z4Xb766qvw8PCATqfDI488grKyMmRnZ2PcuHF45513cN9992HUqFEICAjown8tJBUciZAktLW1oaysDBMmTMCZM2egUChw11134ccff8TevXsRFhaGDz74AOnp6Zg3bx6MRuMNt/P888/jvffew+jRo/HRRx/BYDDg7bff/tVoxMfHB7W1tSgqKsLYsWN/tR1fX19MmzYNw4cPx5dffon3338fqampSE9Px44dOwAA/fv3x6xZsyxPMgcEBGDw4MGYPn06Jk2ahMLCQly8eBF+fn6WwACuPfl8/b9VKpXl6+bmZpSWlgIAhg8fjjlz5qCsrAyNjY2oqKjA0qVLMWLEiE617t+/H5MnT8bLL78MLy8vAIBCoYDJZIJCocCBAwcQEBCAESNGYPLkydiyZQuUSiXuuusu/PDDD2hpaUFjYyOCgoIwYcIEy3YbGxtRVlaGV199Fc899xyKi4sBAOPGjcPLL7+Mzz///DZ/1WuqqqoQEBCAyMhIDB06FA888ABiY2Nx9epVuLu7Y+zYsQwQB8YQIUk4dOgQ6urqkJycjLq6Ohw5cgR/+MMfsHfvXpSXl6N///4Abr/0xIABA7B48WJUV1cjOjoapaWlGDVq1K+mfYBr63Pt2LEDw4YN+9XPNBoNgGsn47a2NnQ8TnX9/j09PTu95/qHVN3c3ODn54fVq1cjOjq60+vMZjNaWlos/25ra0NhYSHi4+MRGBhoGRl1BE3HvjvC5vrQ6aBUKjv9bPPmzUhISMAzzzyD5ubmTrWZTCZMnDgRsbGxyMrKgqenp2WUExcX96ttdxx3Rx2/rOt2srOzodFoMH/+fMu2AODxxx/HwoUL8cUXX1imH8nxcDqLJGHfvn3IycmBh4cHamtrsXz5crz11lt4/fXXERUVBQB44YUXkJSUhL59+1pOZPHx8Vi1ahXkcrllW+fPn4dWq4Wvry+0Wi127Nhxw9VXn3rqKTz55JN3VN/UqVOxYsUKyOVyTJ48+Yav8fb2RkVFBdavX485c+Zg4sSJ+P777zutVQQACxYsQHJyMjw8PNDW1oZ58+ZBp9Ph3XffxdmzZzFy5MhfbdvDwwP33nsvcnNzUVZWhgEDBlh+NnHiRGRlZeHcuXOorKwEAAwcOBDr1q1DdXU1vL290a9fP2zcuBFKpRJ//vOfkZmZCR8fH5hMJsyYMQM5OTlQKBTw9/fvtM9+/fohIyMDV65cwWuvvWZZTflmMjIyoFQqMXjwYADXAnLVqlVQq9X47W9/CwBwdXXFm2++iaCgIBw5cgTNzc0IDg6+g78CSRGfWCeHUVFRgd27d6Ourg5PPvnkDaehpOLixYvIzs7G3LlzO53wiZwNQ4SIiATjNREiIhKMIUJERIIxRIiISDCGCBERCcYQISIiwRgiREQk2P8H0txNDAATxlIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "d=pd.read_csv('data_figure_2.csv')\n",
    "sns.set(context='paper', style='white', palette='gray', rc={'lines.linewidth':.70}, font_scale=.75)\n",
    "sns.barplot(y=d['cparty_code'], x=d['party_minority_pct'], ci='sd', estimator=np.mean, orient='h', order=[22,23,97,59,9,152, 46, 113, 64,63, 73,91])\n",
    "plt.xlabel('Avg. % Minority Candidates on Lists')\n",
    "plt.yticks(np.arange(0, 12, step=1),['SSDP', 'SSDP-GKR', 'NSL', 'LP', 'LZS', 'ZZS', 'JKP',  'Vienotība', 'LRA-LA', 'LRA',\n",
    "                                     'LZP', 'NA'])\n",
    "plt.ylabel(' ')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 3:  Distribution of N Minority Candidates on List Variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/frankthames/.local/lib/python3.8/site-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a=pd.read_csv('data_figure_3.csv')\n",
    "a=count_data['minority_count'].dropna()\n",
    "sns.distplot(a, rug=False, axlabel='N Minority Candidates on Lists')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 4:  First Differences Models 1 and 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#UNZIP MODEL POSTERIORS\n",
    "\n",
    "az_fit_pz_hlm_m1=az.from_netcdf('model_1')\n",
    "az_fit_pz_hlm_m2=az.from_netcdf('model_2')\n",
    "\n",
    "func_dict = {\n",
    "    \"5%\": lambda x: np.percentile(x, 5),\n",
    "    \"50%\": lambda x: np.percentile(x, 50),\n",
    "    \"95%\": lambda x: np.percentile(x, 95),\n",
    "}\n",
    "\n",
    "fd_beta2_m1=az.summary(\n",
    "    az_fit_pz_hlm_m1,\n",
    "    var_names=['fd_beta2'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "fd_beta2_m2=az.summary(\n",
    "    az_fit_pz_hlm_m2,\n",
    "    var_names=['fd_beta2'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "fd_beta2_m1=fd_beta2_m1.reset_index(drop=True)\n",
    "\n",
    "fd_beta2_m2=fd_beta2_m2.reset_index(drop=True)\n",
    "\n",
    "col1={'Model': ['Model 1','Model 1', 'Model 1', 'Model 1','Model 1', 'Model 1', 'Model 1' ], \n",
    "      'Variable':['% Minority Voters', '% University Grads.', '% Population Over 65', '% Unemployment', 'Multi-Region Party (1,0)', \n",
    "                 'Electoral Coaliton (1,0)', 'Left Party (1,0)'] }\n",
    "\n",
    "col2={'Model': ['Model 2','Model 2', 'Model 2', 'Model 2','Model 2', 'Model 2', 'Model 2'], \n",
    "      'Variable':['% Nationalist Vote', '% University Grads.', '% Population Over 65', '% Unemployment',  \n",
    "                  'Multi-Region Party (1,0)', \n",
    "                 'Electoral Coaliton (1,0)', 'Left Party (1,0)'] }\n",
    "var1=pd.DataFrame(col1)\n",
    "var2=pd.DataFrame(col2)\n",
    "fd_beta2_m1=var1.join(fd_beta2_m1)\n",
    "fd_beta2_m2=var2.join(fd_beta2_m2)\n",
    "fd_beta2_m1_m2=fd_beta2_m1.append(fd_beta2_m2).sort_values(['Variable']).reset_index(drop=True)\n",
    "cols=['yaxis']\n",
    "lst=[]\n",
    "start=14\n",
    "\n",
    "for a in range(14):\n",
    "\tlst.append([start])\n",
    "\tstart=start-1\n",
    "order=pd.DataFrame(lst, columns=cols)\n",
    "fd_beta2_m1_m2=fd_beta2_m1_m2.join(order)\n",
    "fd_beta2_m1_m2['yaxis'].replace({12:10, 11:9, 10:8, 9:7, 8:6, 7:5, 6:4, 5:3, 4:12, 3:11}, inplace=True)\n",
    "fd_beta2_m1_m2=fd_beta2_m1_m2.sort_values('yaxis', ascending=False).reset_index(drop=True)\n",
    "\n",
    "sns.set(context='paper', style='white', palette='gray', font_scale=.85, rc={'lines.linewidth':.70})\n",
    "fig=plt.figure()\n",
    "ax=fig.add_subplot(1,1,1)\n",
    "plt.subplots_adjust(bottom=0.1, top=0.9, left=0.25, right=0.75)\n",
    "markers_list=['D', 'o']\n",
    "sns.scatterplot(x= '50%',y= 'yaxis',  data=fd_beta2_m1_m2, hue='Model', style='Model', \n",
    "                markers=markers_list, zorder=5)\n",
    "plt.hlines(y='yaxis', xmin='5%', xmax='95%', data=fd_beta2_m1_m2)\n",
    "plt.ylabel(' '), \n",
    "plt.xlabel('Change in Minority List Candidates')\n",
    "plt.axvline(x=0, linestyle=':')\n",
    "plt.yticks([1,2,3,4,5,6,7,8,9,10,11,12, 13,14], [' ', 'Multi-Region Party (1,0)',  ' ',  'Coalition (1,0)',  \n",
    "                                                 ' ',   'Log % University Gras.',\n",
    "                                                 ' ', 'Log % Unemployment',\n",
    "                                                 ' ', 'Log % Population Over 65', \n",
    "                                                  ' '   , 'Left Party (1,0)', \n",
    "                                                   'Log % Noncitizens', \n",
    "                                                   'Log % Minority Voters' ])\n",
    "plt.legend(loc='lower right')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 5:  Marginal Effect of Left Over Log % Minority Voters, Model 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#UNZIP MODEL POSTERIORS\n",
    "\n",
    "az_fit_pz_hlm_m3=az.from_netcdf('model_3')\n",
    "\n",
    "func_dict = {\n",
    "    \"5%\": lambda x: np.percentile(x, 5),\n",
    "    \"50%\": lambda x: np.percentile(x, 50),\n",
    "    \"95%\": lambda x: np.percentile(x, 95),\n",
    "}\n",
    "\n",
    "\n",
    "marg_effect_m3=az.summary(\n",
    "    az_fit_pz_hlm_m3,\n",
    "    var_names=['marg_effect'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "n_val_mv=34\n",
    "cols=['val_mv']\n",
    "lst=[]\n",
    "start=0\n",
    "\n",
    "for a in range(34):\n",
    "\tlst.append([start])\n",
    "\tstart=start+.1\n",
    "df=pd.DataFrame(lst, columns=cols)\n",
    "\n",
    "marg_effect_m3=marg_effect_m3.reset_index(drop=True)\n",
    "marg_effect_m3=pd.concat([df,marg_effect_m3], axis=1, sort=False)\n",
    "sns.set(context='paper', style='white', palette='gray', font_scale=1, \n",
    "        rc={'lines.linewidth':.70})\n",
    "sns.lineplot(x='val_mv', y='50%', ci=None, data=marg_effect_m3, label='Marginal Effect')\n",
    "plt.fill_between(x='val_mv', y1='5%', y2='95%', data=marg_effect_m3, \n",
    "                 alpha=.25, label='90% Cr.I.')\n",
    "plt.axhline(y=0, linestyle=':')\n",
    "plt.ylabel('N Minority Candidates')\n",
    "plt.xlabel('Log % Minority Voters')\n",
    "plt.legend(loc='best')\n",
    "plt.savefig('model_3.png', dpi=800);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 6:  Marginal Effect of Left Over Log % Non Citizens, Model 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#UNZIP MODEL POSTERIORS\n",
    "\n",
    "az_fit_pz_hlm_m4=az.from_netcdf('model_4')\n",
    "\n",
    "func_dict = {\n",
    "    \"5%\": lambda x: np.percentile(x, 5),\n",
    "    \"50%\": lambda x: np.percentile(x, 50),\n",
    "    \"95%\": lambda x: np.percentile(x, 95),\n",
    "}\n",
    "\n",
    "\n",
    "marg_effect_m4=az.summary(\n",
    "    az_fit_pz_hlm_m4,\n",
    "    var_names=['marg_effect'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "n_val_pnc=29\n",
    "cols=['val_pnc']\n",
    "lst=[]\n",
    "start=1\n",
    "\n",
    "for a in range(29):\n",
    "\tlst.append([start])\n",
    "\tstart=start+1\n",
    "df=pd.DataFrame(lst, columns=cols).round(1)\n",
    "\n",
    "marg_effect_m4=marg_effect_m4.reset_index(drop=True)\n",
    "marg_effect_m4=pd.concat([df,marg_effect_m4], axis=1, sort=False)\n",
    "sns.set(context='paper', style='white', palette='gray', font_scale=1, \n",
    "        rc={'lines.linewidth':.70})\n",
    "sns.lineplot(x='val_pnc', y='50%', ci=None, data=marg_effect_m4,\n",
    "            label='Marginal Effect')\n",
    "plt.fill_between(x='val_pnc', y1='5%', y2='95%', data=marg_effect_m4, \n",
    "                 label='90% Cr.Int.', alpha=.25)\n",
    "plt.axhline(y=0, linestyle=':')\n",
    "plt.ylabel('N Minority Candidates')\n",
    "plt.xlabel('Log % Noncitizens')\n",
    "plt.legend(loc='best')\n",
    "plt.savefig('model_4.png', dpi=800);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 6:  FDs of Model 5 and 6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#UNZIP MODEL POSTERIORS\n",
    "\n",
    "az_fit_hlm_m6=az.from_netcdf('model_5')\n",
    "az_fit_hlm_m7=az.from_netcdf('model_6')\n",
    "\n",
    "func_dict = {\n",
    "    \"5%\": lambda x: np.percentile(x, 5),\n",
    "    \"50%\": lambda x: np.percentile(x, 50),\n",
    "    \"95%\": lambda x: np.percentile(x, 95),\n",
    "}\n",
    "\n",
    "fd_beta_m6=az.summary(\n",
    "    az_fit_hlm_m6,\n",
    "    var_names=['fd_beta'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "\n",
    "fd_beta_m7=az.summary(\n",
    "    az_fit_hlm_m7,\n",
    "    var_names=['fd_beta'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "\n",
    "fd_beta_m6=fd_beta_m6.reset_index(drop=True)\n",
    "\n",
    "fd_beta_m7=fd_beta_m7.reset_index(drop=True)\n",
    "\n",
    "\n",
    "col6={'Model':['Model 5', 'Model 5', 'Model 5', 'Model 5', 'Model 5', 'Model 5',\n",
    "              'Model 5', 'Model 5', 'Model 5', 'Model 5'],\n",
    "    'Variable':['Minority (1,0)', 'Resident (1,0)','Multi-Region Party (1,0)', \n",
    "                 'Electoral Coaliton (1,0)', 'Left Party (1,0)', 'Log % Minority Voters', \n",
    "                  'Log % University Grads.',  'Log % Population Over 65', 'Log % Unemployment', \n",
    "                  'Log % Minority Candidates'], \n",
    "     'Order':[9, 6, 5,    8, 7,4, 3, 2, 1,0],\n",
    "     'Number':[2,2,2,2,2,2,2,2,2,2]}\n",
    "\n",
    "col7={'Model':['Model 6', 'Model 6', 'Model 6', 'Model 6', 'Model 6', 'Model 6',\n",
    "              'Model 6', 'Model 6', 'Model 6', 'Model 6'],\n",
    "    'Variable':['Minority (1,0)', 'Resident (1,0)','Multi-Region Party (1,0)', \n",
    "                 'Electoral Coaliton (1,0)', 'Left Party (1,0)', 'Log % Non Citizens', \n",
    "                  'Log % University Grads.',  'Log % Population Over 65', 'Log % Unemployment', \n",
    "                  'Log % Minority Candidates'],\n",
    "    'Order':[9, 6, 5,    8, 7,4, 3, 2, 1,0],\n",
    "     'Number': [1,1,1,1,1,1,1,1,1, 1]}\n",
    "\n",
    "\n",
    "var6=pd.DataFrame(col6)\n",
    "var7=pd.DataFrame(col7)\n",
    "fd_beta_m6=var6.join(fd_beta_m6)\n",
    "fd_beta_m7=var7.join(fd_beta_m7)\n",
    "fd_beta_m6_m7=fd_beta_m6.append(fd_beta_m7).sort_values(['Variable']).reset_index(drop=True)\n",
    "fd_beta_m6_m7=fd_beta_m6_m7.sort_values(['Order', 'Number']).reset_index(drop=True)\n",
    "\n",
    "\n",
    "cols=['yaxis']\n",
    "lst=[]\n",
    "start=1\n",
    "\n",
    "for a in range(20):\n",
    "\tlst.append([start])\n",
    "\tstart=start+1\n",
    "order=pd.DataFrame(lst, columns=cols)\n",
    "\n",
    "fd_beta_m6_m7=fd_beta_m6_m7.join(order)\n",
    "fd_beta_m6_m7=fd_beta_m6_m7.sort_values('yaxis').reset_index(drop=True)\n",
    "sns.set(context='paper', style='white', palette='gray', font_scale=.85, rc={'lines.linewidth':.70})\n",
    "fig=plt.figure()\n",
    "ax=fig.add_subplot(1,1,1)\n",
    "plt.subplots_adjust(bottom=0.1, top=0.9, left=0.25, right=0.75)\n",
    "markers_list=['D', 'o']\n",
    "sns.scatterplot(x= '50%',y= 'yaxis',  data=fd_beta_m6_m7, hue='Model', style='Model', \n",
    "                markers=markers_list, zorder=5)\n",
    "plt.hlines(y='yaxis', xmin='5%', xmax='95%', data=fd_beta_m6_m7)\n",
    "plt.ylabel(' '), \n",
    "plt.xlabel('Change in Minority List Candidates')\n",
    "plt.axvline(x=0, linestyle=':')\n",
    "plt.yticks([1,2,3,4,5,6,7,8,9,10,11,12, 13,14, 15, 16, 17, 18, 19, 20], [' ', 'Log % Minority Candidates', \n",
    "                                                                        ' ', 'Log % Unemployment ', \n",
    "                                                                        ' ', 'Log % Population Over 65 ', \n",
    "                                                                        ' ', 'Log % University Grad.', \n",
    "                                                                        'Log % Noncitizens ', 'Log % Minority Voters', \n",
    "                                                                        ' ', 'Multi-Region Party (1,0)', \n",
    "                                                                        ' ', 'Resident (1,0)', \n",
    "                                                                        ' ', 'Left Party (1,0)', \n",
    "                                                                        ' ', 'Electoral Coalition (1,0)', \n",
    "                                                                        ' ', 'Minority (1,0)',])\n",
    "\n",
    "plt.legend(loc='best')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 7:  Model 7, Marginal Effect of Left over pct minority voters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#UNZIP MODEL POSTERIORS\n",
    "\n",
    "az_fit_hlm_m8=az.from_netcdf('model_7')\n",
    "\n",
    "func_dict = {\n",
    "    \"5%\": lambda x: np.percentile(x, 5),\n",
    "    \"50%\": lambda x: np.percentile(x, 50),\n",
    "    \"95%\": lambda x: np.percentile(x, 95),\n",
    "}\n",
    "\n",
    "\n",
    "marg_effect_m8=az.summary(\n",
    "    az_fit_hlm_m8,\n",
    "    var_names=['marg_effect'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "n_val_mv=34\n",
    "cols=['val_mv']\n",
    "lst=[]\n",
    "start=0\n",
    "\n",
    "for a in range(34):\n",
    "\tlst.append([start])\n",
    "\tstart=start+.1\n",
    "df=pd.DataFrame(lst, columns=cols)\n",
    "\n",
    "marg_effect_m8=marg_effect_m8.reset_index(drop=True)\n",
    "marg_effect_m8=pd.concat([df,marg_effect_m8], axis=1, sort=False)\n",
    "sns.set(context='paper', style='white', palette='gray', font_scale=1, \n",
    "        rc={'lines.linewidth':.70})\n",
    "sns.lineplot(x='val_mv', y='50%', ci=None, data=marg_effect_m8, label='Marginal Effect')\n",
    "plt.fill_between(x='val_mv', y1='5%', y2='95%', data=marg_effect_m8, \n",
    "                 alpha=.25, label='90% Cr.I.')\n",
    "plt.axhline(y=0, linestyle=':')\n",
    "plt.ylabel('List Position')\n",
    "plt.xlabel('Log % Minority Voters')\n",
    "plt.legend(loc='best')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 8:  Model 8, Marginal Effect of minority over pct non citizens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#UNZIP MODEL POSTERIORS\n",
    "\n",
    "az_fit_hlm_m9=az.from_netcdf('model_8')\n",
    "\n",
    "func_dict = {\n",
    "    \"5%\": lambda x: np.percentile(x, 5),\n",
    "    \"50%\": lambda x: np.percentile(x, 50),\n",
    "    \"95%\": lambda x: np.percentile(x, 95),\n",
    "}\n",
    "\n",
    "\n",
    "marg_effect_m9=az.summary(\n",
    "    az_fit_hlm_m9,\n",
    "    var_names=['marg_effect'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "\n",
    "n_val_na=88\n",
    "cols=['val_na']\n",
    "lst=[]\n",
    "start=0\n",
    "\n",
    "for a in range(88):\n",
    "\tlst.append([start])\n",
    "\tstart=start+1\n",
    "df=pd.DataFrame(lst, columns=cols)\n",
    "\n",
    "marg_effect_m9=marg_effect_m9.reset_index(drop=True)\n",
    "marg_effect_m9=pd.concat([df,marg_effect_m9], axis=1, sort=False)\n",
    "sns.set(context='paper', style='white', palette='gray', font_scale=1, \n",
    "        rc={'lines.linewidth':.70})\n",
    "sns.lineplot(x='val_na', y='50%', ci=None, data=marg_effect_m9,\n",
    "            label='Marginal Effect')\n",
    "plt.fill_between(x='val_na', y1='5%', y2='95%', data=marg_effect_m9, \n",
    "                 label='90% Cr.Int.', alpha=.25)\n",
    "plt.axhline(y=0, linestyle=':')\n",
    "plt.ylabel('List Position')\n",
    "plt.xlabel('Log % Noncitizens')\n",
    "plt.legend(loc='best')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Table 1, Model 9--marg effect Left * Minority"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#UNZIP MODEL POSTERIORS\n",
    "\n",
    "az_fit_hlm_m10=az.from_netcdf('model_9')\n",
    "\n",
    "func_dict = {\n",
    "    \"5%\": lambda x: np.percentile(x, 5),\n",
    "    \"50%\": lambda x: np.percentile(x, 50),\n",
    "    \"95%\": lambda x: np.percentile(x, 95),\n",
    "}\n",
    "\n",
    "marg_effect_m10=az.summary(\n",
    "    az_fit_hlm_m10,\n",
    "    var_names=['marg_effect'],\n",
    "    stat_funcs=func_dict,\n",
    "    extend=False, \n",
    "    round_to=3\n",
    ")\n",
    "marg_effect_m10=marg_effect_m10.reset_index()\n",
    "marg_effect_m10=marg_effect_m10.rename(columns={'index':'Interaction'})\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pz_model_hlm=pystan.StanModel(file='poisson_zip_hlm_m1_m2.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_values():\n",
    "     return  dict (nu=4, sigma_b1=stats.halfnorm.rvs(0, 1), sigma_b2=stats.halfnorm.rvs(0, 1),\n",
    "                   sigma_p1=stats.invgamma.rvs(5,5), sigma_p2=stats.invgamma.rvs(5,5), \n",
    "                   sigma_c1=stats.invgamma.rvs(5,5), sigma_c2=stats.invgamma.rvs(5,5), alpha1=0, \n",
    "                   alpha2=0, eta1=np.random.normal(0,1,size=(k)),  beta2=np.random.normal(0,1,size=(k)), \n",
    "                   ranef_p_1=np.random.normal(0,1, size=(p)), ranef_p_2=np.random.normal(0,1, size=(p)),\n",
    "                  ranef_c_1=np.random.normal(0,1, size=(c)),   ranef_c_2=np.random.normal(0,1, size=(c))   )\n",
    "\n",
    "fit_model_1=pz_model_hlm.sampling(data=data_zip_hlm_m1, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, control=dict(max_treedepth=16))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pz_model_hlm=pystan.StanModel(file='poisson_zip_hlm_m1_m2.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_values():\n",
    "     return  dict (nu=4, sigma_b1=stats.halfnorm.rvs(0, 1), sigma_b2=stats.halfnorm.rvs(0, 1),\n",
    "                   sigma_p1=stats.invgamma.rvs(5,5), sigma_p2=stats.invgamma.rvs(5,5), \n",
    "                   sigma_c1=stats.invgamma.rvs(5,5), sigma_c2=stats.invgamma.rvs(5,5), alpha1=0, \n",
    "                   alpha2=0, eta1=np.random.normal(0,1,size=(k)),  beta2=np.random.normal(0,1,size=(k)), \n",
    "                   ranef_p_1=np.random.normal(0,1, size=(p)), ranef_p_2=np.random.normal(0,1, size=(p)),\n",
    "                  ranef_c_1=np.random.normal(0,1, size=(c)),   ranef_c_2=np.random.normal(0,1, size=(c))   )\n",
    "fit_model_2=pz_model_hlm.sampling(data=data_zip1_hlm_m2, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, control=dict(max_treedepth=16))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pz_model_hlm_m3=pystan.StanModel(file='poisson_zip_hlm_m3.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_values():\n",
    "     return  dict (nu=4, sigma_b1=stats.halfnorm.rvs(0, 1), sigma_b2=stats.halfnorm.rvs(0, 1),\n",
    "                   sigma_p1=stats.invgamma.rvs(5,5), sigma_p2=stats.invgamma.rvs(5,5), \n",
    "                   sigma_c1=stats.invgamma.rvs(5,5), sigma_c2=stats.invgamma.rvs(5,5), alpha1=0, \n",
    "                   alpha2=0, eta1=np.random.normal(0,1,size=(k)),  beta2=np.random.normal(0,1,size=(k)), \n",
    "                   ranef_p_1=np.random.normal(0,1, size=(p)), ranef_p_2=np.random.normal(0,1, size=(p)),\n",
    "                  ranef_c_1=np.random.normal(0,1, size=(c)),   ranef_c_2=np.random.normal(0,1, size=(c))   )\n",
    "\n",
    "fit_mdoel_3=pz_model_hlm_m3.sampling(data=data_zip1_hlm_m3, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, control=dict(max_treedepth=15))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model 4, Hypothesis 4b--IV:Left*Non Citizens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL anon_model_259a47c6e7212985853010e890fffecb NOW.\n"
     ]
    }
   ],
   "source": [
    "pz_model_hlm_m4=pystan.StanModel(file='poisson_zip_hlm_m4.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_values():\n",
    "     return  dict (nu=4, sigma_b1=stats.halfnorm.rvs(0, 1), sigma_b2=stats.halfnorm.rvs(0, 1),\n",
    "                   sigma_p1=stats.invgamma.rvs(5,5), sigma_p2=stats.invgamma.rvs(5,5), \n",
    "                   sigma_c1=stats.invgamma.rvs(5,5), sigma_c2=stats.invgamma.rvs(5,5), alpha1=0, \n",
    "                   alpha2=0, eta1=np.random.normal(0,1,size=(k)),  beta2=np.random.normal(0,1,size=(k)), \n",
    "                   ranef_p_1=np.random.normal(0,1, size=(p)), ranef_p_2=np.random.normal(0,1, size=(p)),\n",
    "                  ranef_c_1=np.random.normal(0,1, size=(c)),   ranef_c_2=np.random.normal(0,1, size=(c))   )\n",
    "\n",
    "fit_model_4=pz_model_hlm_m4.sampling(data=data_zip1_hlm_m4, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, control=dict(max_treedepth=14))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL anon_model_08ca4b2cfa14517b4fc4e272386654bd NOW.\n"
     ]
    }
   ],
   "source": [
    "hlm_5_6=pystan.StanModel(file='hlm_m5_6.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_values():\n",
    "    return dict(intercept=np.mean(y_list), beta=np.random.normal(0,1,size=k_list), sigma_b=1)\n",
    "\n",
    "fit_model_5=hlm_m5_m6.sampling(data=data_hlm_m6, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, init=init_values,  control=dict(max_treedepth=15))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "hlm_5_6=pystan.StanModel(file='hlm_m5_6.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_values():\n",
    "    return dict(intercept=np.mean(y_list), beta=np.random.normal(0,1,size=k_list), sigma_b=1)\n",
    "\n",
    "fit_model_6=hlm_5_6.sampling(data=data_hlm_m7, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, init=init_values,  control=dict(max_treedepth=14))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL anon_model_17990f43edaea0159e58f05e149e9bb7 NOW.\n"
     ]
    }
   ],
   "source": [
    "hlm_7_8=pystan.StanModel(file='hlm_m7_m8.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "def init_values():\n",
    "    return dict(intercept=np.mean(y_list), beta=np.random.normal(0,1,size=k_list), sigma_b=1)\n",
    "\n",
    "fit_model_7=hlm_m7_m9.sampling(data=data_hlm_m8, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, init=init_values,  control=dict(max_treedepth=15))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL anon_model_17990f43edaea0159e58f05e149e9bb7 NOW.\n"
     ]
    }
   ],
   "source": [
    "hlm_7_8=pystan.StanModel(file='hlm_m7_m8.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def init_values():\n",
    "    return dict(intercept=np.mean(y_list), beta=np.random.normal(0,1,size=k_list), sigma_b=1)\n",
    "\n",
    "fit_model_8=hlm_m7_m8.sampling(data=data_hlm_m9, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, init=init_values,  control=dict(max_treedepth=14))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model 9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL anon_model_28c8664d92b0199db7c4f747b3f1cf4c NOW.\n"
     ]
    }
   ],
   "source": [
    "hlm_9=pystan.StanModel(file='hlm_m9.stan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "def init_values():\n",
    "    return dict(intercept=np.mean(y_list), beta=np.random.normal(0,1,size=k_list), sigma_b=1)\n",
    "\n",
    "fit_model_9=hlm_9.sampling(data=data_hlm_m10, warmup=10000, iter=20000, \n",
    "                                 chains=4, thin=10, init=init_values,  control=dict(max_treedepth=14))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}
